Copd Monitoring

1. A system for planning resources for a plurality of chronic obstructive pulmonary disease, COPD, remote patient monitoring, RPM, programs, each RPM program being tailored to a different level of disease acuity, the system comprising: a processor adapted to: receive assessment data relating to the severity of COPD symptoms in respect of a set of patients currently following the RPM programs and/or in respect of a set of patients who are prospective eligible candidates for the RPM programs, wherein the assessment data comprises a severity measure for one or more of: a COPD assessment test trend; a COPD assessment test impact level; a measure of a number of exacerbation events; a disease severity measure; and an oxygen saturation level; calculate a score which combines the plurality of severity measures of the assessment data and allocate the patient to a RPM program based on at least the score; for each patient, predict which RPM program is most likely to be appropriate for the patient at a future time point by applying a prediction algorithm to the assessment data to predict a required future distribution of patients between different RPM programs, wherein the predication algorithm comprises a function, Pr ⁡ ( Y i = K - 1 ) = e β K - 1 t * x i 1 + ∑ K = 1 K - 1 ⁢ e β k t * x i , where Pr is the predicted outcome probability, and K is the number of RPM programs, and β(k) is a vector of coefficients of input features associated to outcome k, wherein the input features are based on the assessment data, and a function, Pr n p = max p Pr ⁡ ( Y n = p ) , wherein for each patient n, the program allocation probability for program p is given; determine a required capacity for each RPM program at said future time point based on the program allocation probability; and output data identifying the required resources for each RPM program at said future time point.

2. The system of claim 1, wherein the processor is further adapted to: determine a required capacity for each RPM program at a near term future time point based on the program allocation probability; and determine a required capacity for each RPM program at a long term future time point based on the program allocation probability.

3. The system of claim 1, wherein model coefficients βit are trained for different time horizons.

4. The system of claim 3, wherein the processor is adapted to derive a probability for each patient using a multinomial logistic regression.

5. The system of claim 1, wherein the processor is adapted to derive a cost budget for: each RPM program; or each of a set of resources which together constitute the RPM programs.

6. The system of claim 1, wherein the processor is adapted to determine a time-to availability for the required resources for each RPM program at said future time point.

7. The system of claim 1, wherein the plurality of RPM programs comprises a first RPM program for low acuity, a second RPM program for medium acuity and a third RPM program for high acuity.

8. The system of claim 1, wherein the severity measure for the COPD assessment test trend is based on a measure of a gradient of a line of best fit to a COPD assessment test value over time.

9. The system of claim 1, wherein the severity measure for the COPD assessment test impact level comprises a value for a low impact, a medium impact, a high impact and a very high impact.

10. The system of claim 1, wherein the severity measure for the oxygen saturation level comprises value for a very low level, a low level and a normal level.

11. The system of claim 1, wherein the processor is further adapted to receive an identification of a current RPM program to which a patient is currently allocated, wherein the current RPM program is allocated a base severity measure, and wherein each severity measure comprises a numeric value, wherein the numeric value depends on the current RPM program.

12. A computer-implemented method for planning resources for a plurality of chronic obstructive pulmonary disease, COPD, remote patient monitoring, RPM, programs, each RPM program being tailored to a different level of disease acuity, the method comprising: receiving assessment data relating to the severity of COPD symptoms in respect of a set of patients currently following one of the RPM programs, wherein the assessment data comprises a severity measure for one or more of: a COPD assessment test trend; a COPD assessment test impact level; a measure of a number of exacerbation events; a disease severity measure; and an oxygen saturation level; calculating a score which combines the plurality of severity measures of the assessment data and allocating the patient to a RPM program based on at least the score; for each patient, predicting which RPM program is most likely to be appropriate for the patient at a future time point by applying a prediction algorithm to the assessment data to predict a required future distribution of patients between different RPM programs, wherein the predication algorithm comprises a function, Pr ⁡ ( Y i = K - 1 ) = e β K - 1 t * X i 1 + ∑ K = 1 K - 1 e β k t * X i , where Pr is the predicted outcome probability, and K is the number of RPM programs, and β(k) is a vector of coefficients of input features associated to outcome k, wherein the input features are based on the assessment data, and a function, Pr n p = max p Pr ⁡ ( Y n = p ) , wherein for each patient n, the program allocation probability for program p is given; determining a required capacity for each RPM program at said future time point based on the program allocation probability; and outputting data identifying the required resources for each RPM program.

13. A computer program comprising computer program code means which is adapted, when said computer program is run on a computer, to implement the method of claim 12.